Biological membranes surround all cells and mediate their interactions with the outside world. Depending on the biological context, membrane proteins act as receptors, enzymes, channels, transporters, structural proteins and cell adhesion molecules and, as such, contribute to a wide variety of essential cellular functions. Structural information for membrane proteins is relatively scarce, despite the fact that they represent the target of 60% of therapeutic drugs sold in the United States. We propose to establish a pipeline for determining membrane protein structures by electron crystallography, which is the application of cryo-electron microscopy to two-dimensional crystals. To prepare such crystals, the membrane bilayer is reconstituted with a high density of purified membrane proteins, thus providing a native membrane environment and fewer crystallization constraints for the constituent proteins. Electron crystallography has an established track record in producing structures at both atomic and intermediate resolutions and represents a valuable alternative to X-ray crystallography and NMR spectroscopy, which are generally constrained to studying detergent-solubilized species. To date, electron crystallography remains a low-throughput operation, which has significantly reduced its contribution to membrane protein biology. We have developed some tools to overcome the bottlenecks in screening crystallization conditions, which we seek to expand in the current application. Specifically, we propose further developments for crystallization on a 96-well format by implementing a microfluidic device for dialysis that minimizes sample volumes and by using cyclodextrins to control detergent removal rates in an effort to optimize crystal quality. By studying a wide range of different target proteins, we will empirically establish factors that are most critical to influencing the crystallization process and develop a set of conditions that are generally effective for screening new proteins with unknown crystallization behaviors. We will continue developing our methods for robotic imaging of crystallization screens. The process of preparing samples and imaging them within the electron microscope currently represents the most significant bottleneck limiting the number of conditions that can be explored. We have built a robot for sample insertion and have interfaced it with automated image acquisition software, but we propose to add shape recognition to this software to maximize its efficiency. We will integrate the resulting images within an established LIMS database for keeping track of the structure determination pipeline and will implement shape recognition to enable automated assignment of scores to crystallization trials. Finally, we propose to develop an application for high resolution data collection from well-ordered crystals, thus facilitating the optimization of crystal size and order and, ultimately, acquisition of the data required for structure determination. We are convinced that by applying high- throughput methods to 2D crystallization and image acquisition, electron crystallography can make a substantial contribution to our understanding of membrane protein biology.